Ancestral Origin of the First Indian Families with Myotonic Dystrophy Type 2

INTRODUCTION

Myotonic dystrophy type 2 (DM2) is a progressive autosomal dominant multisystem disease with muscle dysfunction as a main feature, including proximal muscle weakness, myotonia and pain [1]. In addition, cataracts, cardiac conduction defects, endocrine disorders, gastrointestinal involvement, autoimmune diseases, cerebral involvement, and hearing impairment are common in patients with DM2 [1–10].

DM2 is caused by the expansion of a CCTG repeat in intron 1 of the CCHC-Type Zinc Finger Nucleic Acid Binding Protein (CNBP) gene located on chromosome 3q21.3 [11]. The estimated age of the DM2 mutation in the European haplotype is 4000 –11,000 years [12].

Previous studies show that patients with DM2 from European descent share a common founder as the DM2 mutation in these patients share identical haplotypes. In all these patients, the region in and around the CNBP gene can be classified into the same disease linked haplotypes based on repeat lengths (Fig. 1) [12–14]. Liquori et al. found the majority of the 70 European families from Northern European descent carried one of the three consensus haplotypes and 1 family from Afghanistan with a haplotype with great similarity to the core haplotype [13]. Bachinski et al. also found one single founder mutation in 17 unrelated DM2 pedigrees from both Northern and Southern Europe [12]. Coenen et al. reported 14 European families and 1 Moroccan family with a haplotype similar to this European or Caucasian haplotype [14]. This suggests that the CCTG repeat expansion in CNBP in these patients has the same ancestral origin. However, Saito et al. reported a Japanese patient with DM2 carrying a haplotype different from the common European haplotype [15]. These studies indicate that the CCTG repeat originates from at least two separate founders.

Fig. 1

Previous reported haplotypes in literature. Reported marker lengths of Liquori et al. and Saito et al. are recalculated to make them comparable to our lengths and the lengths previously reported by Coenen et al. [13–15]. NA: not analysed.

Apart from the above-mentioned studies, there is a general lack of research in patients with DM2 originating from non-European countries. This study was set out to determine whether patients with DM2 originating from European and non-European countries carry the same European founder haplotype by investigating the haplotypes of patients with ancestries that have not been examined earlier.

MATERIALS AND METHODS

RESULTS

Sixty-seven patients with DM2 were identified by searching the Dutch neuromuscular database. Two patients were excluded because they were younger than 18 years old. Two registered patients were found to be deceased and 4 patients had DNA that was not available for research; leaving DNA of 59 DM2 patients from 29 unrelated families for analysis, including 14 probands (patient 1, 7, 10, 22, 23, 25, 26, 41, 52, 53, 55–59).

The mean age of genetically confirmed diagnosis was 47.2 years. Twenty-three families were from European descent. Six families (2 families from each country) ethnically originated from non-European countries, namely India, Suriname and Morocco (Table 1).

Haplotype analysis

Results of the haplotype analysis are shown in Figs 2 and 3. SNPs rs1384313, rs4303883 and CGAP_886192 are not included as these did not show genetic variation in our study population. Most of the identified STR marker lengths are in the same range as those reported earlier [13, 14]. In 19 of the 59 patients, only one haplotype with high probability was constructed by Phase (≥0.99). For the other 40 patients, multiple haplotypes were identified. An overview of all haplotypes constructed using Phase is presented in Supplemental File 2. All family members were assigned to the haplotype linked to the DM2 causing CNBP repeat expansion based on the family structure.

Fig. 2

Haplotypes of 59 DM2 patients. (both color and black/white version):NA: not analysed. The length of the analysed markers is expressed in base pairs. Haplotypes are based on earlier reported haplotypes [13–15]. Repeat lengths as reported by Liquori et al. are shown in the first row [13]. The families related to earlier reported patients by Coenen et al. are shown in italic [14]. Patients have been assigned to a haplotype based on the observed repeat lengths. The repeat lengths of the disease-linked alleles are shown in bold. For some individual patients, it was not possible to determine this with certainty. Genotypes shared among the different haplotypes are shaded. ^* Recalculated reported marker lengths of previous studies to make them comparable to our lengths.

Fig. 3

Geographic distribution of the distinct DM2 haplotypes. (color version): Patients with DM2 throughout the world as reported in literature (Canada, USA, Suriname, UK, Norway, Sweden, Finland, The Netherlands, Belgium, France, Spain, Portugal, Germany, Switzerland, Italy, Malta, Poland, Czech Republic, Slovakia, Ukraine, Serbia, Greece, Morocco, Algeria, Lebanon, Afghanistan, India, Sri Lanka and Japan) [11, 12, 14, 15, 24, 25] Orange: European haplotype. Blue: Japanese haplotype. Green: Indian haplotype. Yellow: haplotype does not exactly fit with European haplotype. Dark grey: reported patients with DM2, but no haplotyping study is done. (black/white version): Patients with DM2 throughout the world as reported in literature (Canada, USA, Suriname, UK, Norway, Sweden, Finland, The Netherlands, Belgium, France, Spain, Portugal, Germany, Switzerland, Italy, Malta, Poland, Czech Republic, Slovakia, Ukraine, Serbia, Greece, Morocco, Algeria, Lebanon, Afghanistan, India, Sri Lanka and Japan) [11, 12, 14, 15, 24, 25] Striped pattern: European haplotype. Checkered pattern: Japanese haplotype. Dotted pattern: Indian haplotype. Dark grey: haplotype does not exactly fit with European haplotype. Black: reported patients with DM2, but no haplotyping study is done.

Comparison of our data to the previously reported data showed that 41 patients could be directly assigned to the haplotypes described by Liquori et al. [13]. Though, in 3 patients (1, 26 and 27) the exact haplotype is not clear. Patient 1 fits with haplotype B based on all markers, except for the CL3N59 marker closest to the CCTG repeat expansion in CNBP. Patient 26 fits within haplotype A based on the markers closest to the DM2 repeat (CL3N59 and CL3N117), but the lengths of the other markers are in line with haplotype group C. For patient 27, the length of the markers closest to the CNBP repeat expansion (CL3N59 and CL3N117) fits with haplotype C, but the other markers are more difficult to assign to a haplotype.

Outside the 41 patients assigned to the earlier reported Caucasian haplotypes, 18 patients did not fit well with these haplotypes [13]. For 6 patients (54–59) the lengths of the markers closest to the repeat were not comparable with previously reported lengths, however some of the markers located more distant from the repeat expansion are in line with the previously reported haplotypes. Thus, based on our results we cannot assign these patients to one of the previously reported haplotypes. The 3 probands from Portugal and Morocco (patient 57, 58 and 59) have a haplotype comparable with the Moroccan patients (family XV) reported by Coenen et al. [14]. The probability of the haplotypes of these patients was low for patient 57 (0.35) and higher for the other 2 individuals (>0.83).

Eleven patients were related to 6 families earlier reported by Coenen et al. (family IV, VII, XI-XIV) [14]. In our results, we used the same numbering for these families as reported earlier to compare the haplotype of our patients with their relatives. The length of the repeat markers of our 11 patients are the same as those of their earlier reported relatives.

Four families (XXXI, XXXII, XXXIII and XXXIV) from India and Suriname showed a haplotype different from the common haplotypes, with six markers shared within family XXXI, XXXII and XXXIII over a 661-kb interval. The other patient of Indian descent (patient 53, family XXXIV) shares three markers with these families.

The SNP rs1871922, located 11kb telomeric from the CCTG expansion in intron 1 of the CNBP gene, was analysed in 59 patients. All patients carried at least one C allele. All except for patient 55 and 56, confirmed that the expansion is linked to the C allele. For patient 55 and 56 it remained unclear whether the DM2 causing CNBP repeat expansion is linked to the C or A allele of the SNP. However, as both patients are from Caucasian origin it is highly likely that the expansion is linked to the C allele.

DISCUSSION

In this study, we suggest that the DM2 linked repeat expansion in CNBP may originate from more founders than previously had been thought. Through our research with the Indian and Suriname families, a new haplotype may have been identified. The lengths of the markers of these families do not show a high overlap with either the European or Japanese haplotypes [13–15].

The similarity between the haplotype of these 4 families (Indian and Suriname families XXXI-XXXIV) suggests that the Surinamese patients are connected to the Indian family XXXI. Indeed, it is highly plausible that these are from Indian descent, as the Indians form 27%of the population of Suriname. The Indians in Suriname are descendants of 19th-century contract workers from British India [19]. A previous study established 1 Afghan family carrying a haplotype strikingly similar to the core region of haplotype A. This DM2 linked CNBP repeat expansion may have been introduced when the ancient Aryan tribes of Indo-European extraction settled in ancient Afghanistan [13]. Based on the migration of the Indo-Europeans, it is likely that patients from North Indian descent have similarity to this Afghan haplotype and therefore to haplotype A.

Genetic studies have shown that most Indian groups are descendants of two major genetically divergent populations [20]. The first, Ancestral North Indians, are (genetically) related to Central Asians, Middle Easterners, Caucasians and Europeans (West Eurasians). Ancestral South Indians are restricted to South Asia and are not closely related to groups outside their region, but to indigenous Andaman Islanders. Both groups are also distinct from East Asians (including Japan).

The 4 families included in this study from Indian descent do not have haplotypes similar to the Afghan patients or European patients. This suggests that the haplotype of the Indian patients differs from that of the European haplotype, and therefore possibly originate from Ancestral South India. Unfortunately, exact data regarding the geographic origin of the family of the Indian patients is missing.

In addition, genotype analysis of SNP rs1871922 showed that all patients, including the Indian families, carry at least one C allele. Thus, it might be possible that the Indian families carry a short common haplotype between the markers CL3N59 and CL3N117. Unfortunately, the three additional SNPs (located close to the repeat expansion) did not show variation in any of the patients. We specifically selected all SNPs from literature and databases with high frequency differences between populations. Based on these analyses, we cannot conclude, with certainty, that the Indian families carry a non-Caucasian haplotype as in the case of the reported Japanese founder haplotype (associated with the A allele) [12, 15, 21]. However, based on the previous described migration history, an absence of an A allele and different lengths of the other analysed markers, we propose that the DM2 linked repeat expansion in Indian patients originates from an ancestral origin different from the European and Japanese families.

The similarity between the patients with Indian ancestries and the previously reported Japanese patient is the length of the CL3N117 and CL3N119 markers, which is respectively 245 and 226 bp. The 226 bp allele has a frequency of 10%in Caucasians, in Asia the prevalence is probably equal or higher [13].

We included 14 patients with DM2 that did not have diagnosed family members with DM2 in The Netherlands. In those cases, it is more difficult to determine which alleles are linked with DM2. The 3 individuals (patient 54–56) that do not fit within a European haplotype all have at least 1 parent originating from outside The Netherlands. Patient 54 has a mother from Austria and patients 55 and 56 have a Belgian and Greek parent respectively. These differences in ancestry may explain the wider variations in the length of the analysed STR repeat markers. It would be necessary to analyse relatives of these patients to pinpoint the haplotype linked to DM2 in these patients.

The patients from Portugal (patient 57) and Morocco (patient 58 and 59) have marker lengths comparable with previously reported Moroccan patients [14]. These patients may have another haplotype, or at least variations of the major Caucasian haplotypes, due to potential recombination and microsatellite instability events in flanking regions. Their haplotype has probably arisen from the C haplotype, as the 222 kb region surrounding the CNBP gene of these patients is comparable with this haplotype.

Our study included patients originating from European countries that have not been examined before, namely Ukraine, Belgium, Greece and Portugal. We show that DM2 patients from Ukraine and Belgium carry the European common haplotype and that patients originating from Greece, Portugal and Morocco may have a variant of this haplotype (Figs 2 and 3). These results are in line with previous reports showing that patients from European ancestry carry a common haplotype [12, 13]. Further research should be undertaken to obtain more insight in the worldwide haplotypes of DM2 as data from other regions are lacking [15].

In addition, it would be interesting to investigate whether these haplotypes differ in phenotypic presentation and severity of DM2. For example, research regarding the presence of additional symptoms such as cataracts, cardiac conduction defects and autoimmune diseases in DM2 patients with different haplotypes would be worthwhile to investigate whether these are linked to specific haplotypes.

The mean age of genetically confirmed diagnosis of our study population corresponds to literature with a mean onset between the third and fourth decade [1, 22, 23].

Based on the haplotype studies performed thus far, including this study, we can conclude that the European haplotype is associated with a 241 bp length of the CL3N117 marker, as all tested European patients carry this allele [13, 14]. This repeat length occurs at a frequency of 12%in a healthy Caucasian population in which a length of 245 bp is much more prevalent (82%) [13]. Besides, the DM2 region appears to be highly conserved allele [13]. This all indicates that the European founder CNBP repeat expansion developed on the less frequent 241 bp length. In the Japanese patient, the CL3N117 marker length associated with DM2 was 245 bp. In the Indian haplotype, it is most likely that the CNBP repeat expansion is also linked to the repeat length of 245 bp of CL3N117, as this was the haplotype constructed with highest probability for all members of family XXXI. For 4 patients (43, 45, 47 and 48) the probability was even 1.0. More research in patients with DM2 from India is necessary to validate this observation. Research regarding the most common haplotypes in India would also be interesting and informative.

Our study describes the first Indian patients with DM2, all currently living and diagnosed in The Netherlands. DM2 is probably (significantly) underdiagnosed in Indians. Our research describes 4 different families with a DM2 causing CNBP repeat expansion just inside The Netherlands. The fact that India has a population twice as big as Europe, genetic testing for DM2 in Indian population might indeed show that DM2 is underdiagnosed in this population.

In conclusion, our data shows that patients from Indian descent may carry a haplotype different from the haplotypes identified before. This might suggest a third founder in the etiology of myotonic dystrophy type 2. We also show that most European patients with DM2 carry the common founder haplotype. Further studies in European and especially non-European and Asian DM2 patients are required to confirm our findings.